986 J. Phycol. 36, 986–1011 (2000) REVIEW TRACKING LONG-TERM CHANGES IN CLIMATE USING ALGAL INDICATORS IN LAKE SEDIMENTS 1 John P. Smol 2 and Brian F. Cumming Paleoecological Environmental Assessment and Research Laboratory (PEARL), Department of Biology, Queen’s University, Kingston, Ontario, K7L 3N6, Canada Interest in climate change research has taken on new relevance with the realization that human activi- ties, such as the accelerated release of the so-called greenhouse gases, may be altering the thermal prop- erties of our atmosphere. Important social, economic, and scientific questions include the following. Is cli- mate changing? If so, can these changes be related to human activities? Are episodes of extreme weather, such as droughts or hurricanes, increasing in fre- quency? Long-term meteorological data, on broad spatial and temporal scales, are needed to answer these questions. Unfortunately, such data were never gath- ered; therefore, indirect proxy methods must be used to infer past climatic trends. A relatively untapped source of paleoclimate data is based on hindcasting past climatic trends using the environmental optima and tolerances of algae (especially diatoms) preserved in lake sediment profiles. Paleophycologists have used two approaches. Although still controversial, attempts have been made to directly infer climatic variables, such as temperature, from past algal assemblages. The main assumption with these types of analyses is that species composition is either directly related to temperature or that algal assemblages are related to some variable linearly related to temperature. The second more commonly used approach is to infer a limnological variable (e.g. water chemistry, lake ice cover, etc.) that is related to climate. Although pale- olimnological approaches are broadly similar across climatic regions, the environmental gradients that paleophycologists track can be very different. For example, climatic inferences in polar regions have focused on past lake ice conditions, whereas in lakes near arctic treeline ecotones, paleophycologists have developed methods to infer past lakewater-dissolved organic carbon, because this variable has been linked to the density of coniferous trees in a drainage basin. In closed-basin lakes in arid and semiarid regions, past lakewater salinity, which can be robustly recon- structed from fossil algal assemblages, is closely tied to the balance of evaporation and precipitation (i.e. drought frequency). Some recent examples of paleo- phycolgical work include the documentation of strik- ing environmental changes in high arctic environments in the 19th century believed to be related to climate warming. Meanwhile, diatom-based reconstructions of salinity (e.g. the Great Plains of North America and Africa) have revealed prolonged periods of droughts over the last few millennia that have greatly exceeded those recorded during recent times. Marked climatic variability that is outside the range captured by the instrumental record has a strong bearing on sustainability of human societies. Only with a long- term perspective can we understand natural climatic variability and the potential influences of human activities on climate and thereby increase our ability to understand future climate. Key index words: chrysophytes; climate change; dia- toms; environmental change; fossils; lakes; paleolim- nology; paleoclimatology; pigments; sediments; sto- matocysts Abbreviations: yr B.P., years before present; CCA, canonical correspondence analysis; DIC, dissolved inorganic carbon; DOC, dissolved organic carbon; ELA, Experimental Lakes Area; GCM, general circu- lation model Meteorological data from most North American lo- cations have only been collected in a consistent and reproducible fashion for about the last 100 years. Cli- mate records for many European cities are only slightly longer, with the longest continuous readings extending back to the early 1700s (Hulme 1994). Me- teorological records are even more sparse, or simply nonexistent, for many climatically sensitive areas (e.g. the Arctic and Antarctic, alpine regions, arid regions). Without long-term climate data, however, it is not pos- sible to answer many of the critical questions cur- rently being posed in a world concerned with climatic change, especially of issues related to possible green- house gas-enhanced warming. Typical questions in- clude the following. Is climate changing, or are recent changes simply part of a long-term trend, unrelated to human activities? Are the recent changes in weather patterns related to greenhouse gas emissions, or is this just part of natural variability? Are droughts and other examples of extreme weather increasing in fre- quency, magnitude, and duration, or are they compa- rable with similar events that have occurred in the past? These questions cannot be answered without a long-term perspective that is set at a scale of centuries and millennia, not simply years and decades. Unfortu- 1 Received 29 March 2000. Accepted 1 September 2000. 2 Author for correspondence: e-mail smolj@biology.queensu.ca.